CpG island methylator phenotype (CIMP) is a unique epigenetic phenomenon observed in cancer. Tumors possessing a CIMP usually display concurrent hypermethylation of multiple promoter CpG islands. The causes of CIMP are unknown. An issue of great importance is whether CIMPs are the result of a systematic deregulation of signal transduction in cancer cells. Specifically, our preliminary studies suggest that ERa signaling disruption contributes, in part, to the development of CIMPs in breast cancer. We hypothesize that this disturbance results in altered transcription profiles of ERa-responsive targets, some of which are down- regulated via epigenetic mechanisms. Concurrent hypermethylation of these responsive targets is a hallmark (or molecular relic) of this transcriptional silencing and creates specific patterns, which are epigenetically depicted as CIMPs. In this revised application, global and gene-specific methylation analysis will be conducted in primary tumors, adjacent normals, and unrelated normal mammoplasty samples. A novel computational algorithm, called Heritable Clustering, will be used to generate a phylogenetic model that simulates the history of methylation alteration from pre-neoplastic lesions to hyperplasia to ductal carcinoma in situ to invasive cancers. Different CIMPs profiles generated by Heritable Clustering can represent different stages of breast cancer development. Specific attention will be made to find ERa-related CIMPs that predict patients'intrinsic resistance to endocrine treatment. Genetic and biochemical studies will solidify an important role of ERa co-regulators in contributing to the mechanisms of endocrine resistance. As a result of crosstalk with other signaling pathways, ERa co-regulators can be abnormally expressed in breast cancer cells. This may result in epigenetic silencing of a subset of ERa-responsive targets, which will be validated by quantitative assays and bisulfite sequencing in "resistance-prone" primary tumors. The present study has a practical ramification for clinical management of hormone receptor-positive tumors. Today, the paucity of molecular information to predict the outcome of endocrine treatment in patients brings urgency for identifying new types of biomarkers. In this regard, promoter hypermethylation related to ERa signaling is an untapped resource of predictors for the resistance disease. This type of epigenetic studies will be useful for identifying patients with resistant-CIMPs for alternative therapies in the future.